A method and system for automated inspection of drone nests

Abstract

This invention relates to the field of navigation planning technology, and discloses a method and system for automated UAV nest inspection path planning. The method includes: fusing multi-source data of the target inspection area to obtain spatial measurement environment data; analyzing the airspace constraints in the spatial measurement environment data and combining them with the inspection task requirements of the target inspection area to determine navigation path measurement constraints; performing gridded measurement on the target inspection area to identify candidate waypoints; analyzing the spatial relationships of the candidate waypoints based on spatial location information to construct an initial navigation network; performing multi-constraint evaluation on the initial navigation network to obtain the constraint weights corresponding to the candidate waypoints, and integrating the candidate waypoints with constraint weights higher than a preset weight threshold into a globally optimal waypoint set; and making traversable connections between adjacent waypoints in the globally optimal waypoint set to obtain the navigation path. This invention can improve the efficiency of automated UAV nest inspection path planning.

Description

Technical Field

[0001] This invention relates to the field of navigation planning technology, and in particular to a method and system for planning the path of an unmanned aerial vehicle (UAV) nest for automated inspection. Background Technology

[0002] In existing UAV automated inspection nest path planning technologies, there is a lack of efficient spatial registration and structured integration capabilities for multi-source heterogeneous data of the target inspection area. This results in insufficient completeness and accuracy of the generated spatial measurement environment data, making it difficult to accurately reflect the terrain elevation distribution and the spatial outline of static obstacles. At the same time, when analyzing airspace constraints and determining navigation path measurement constraints in conjunction with inspection task requirements, reasonable terrain safety buffers and obstacle safety buffers are not fully constructed, and the path rule descriptions cannot be fully integrated, resulting in deviations or omissions in the constraint basis of path planning.

[0003] When performing gridded measurements based on constraints, the scientific rigor of defining the grid operation range and selecting feasible waypoints is insufficient, resulting in low accuracy of candidate waypoints. During the initial navigation network construction, the tracking of spatial proximity relationships between candidate waypoints and the analysis of the reachability of adjacent candidate pairs are incomplete, and the rationality of network nodes and connecting edges is lacking. Furthermore, the multi-constraint evaluation dimension is singular, making it impossible to scientifically weight and fuse candidate waypoint constraint weights, thus hindering the selection of the globally optimal waypoint set. The resulting navigation paths often suffer from impassability and poor smoothness, severely impacting the reliability and efficiency of inspection operations. Therefore, how to systematically optimize and improve the accuracy and effectiveness of UAV automated inspection nest path planning has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a method and system for automated inspection of drone nests to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for automated inspection of drone nests, comprising:

[0006] S1. Perform multi-source data fusion on the target inspection area to obtain the spatial measurement environment data of the target inspection area;

[0007] S2. Analyze the airspace constraints in the space measurement environment data and combine them with the inspection task requirements of the target inspection area to determine the navigation path measurement constraints of the target inspection area.

[0008] S3. Based on the navigation path measurement constraints, perform gridded measurement on the target inspection area to identify candidate waypoints for the target inspection area;

[0009] S4. Based on the spatial location information of the candidate waypoints, analyze the spatial relationships of the candidate waypoints to construct an initial navigation network for the target inspection area;

[0010] S5. Perform multi-constraint evaluation on the initial navigation network to obtain the constraint weights corresponding to the candidate waypoints, and integrate the candidate waypoints with constraint weights higher than a preset weight threshold into the global optimal waypoint set of the target inspection area.

[0011] S6. Connect adjacent waypoints in the global optimal waypoint set to obtain the navigation path of the target inspection area.

[0012] In a preferred embodiment, the step of fusing multi-source data of the target inspection area to obtain spatial measurement environment data of the target inspection area includes:

[0013] Geographic information data, remote sensing image data, airspace control data and inspection task parameters of the target inspection area are acquired to obtain multi-source heterogeneous data of the target inspection area;

[0014] By combining the spatial location information of the multi-source heterogeneous data, spatial registration is performed on the multi-source heterogeneous data to obtain the registered multi-source data.

[0015] The registered multi-source data is structured and integrated to obtain the spatial measurement environment data of the target inspection area.

[0016] In a preferred embodiment, the step of parsing the airspace constraints in the spatial measurement environment data and combining them with the inspection task requirements of the target inspection area to determine the navigation path measurement constraints of the target inspection area includes:

[0017] Spatial feature deconstruction is performed on the spatial measurement environment data to obtain the terrain elevation distribution information and static obstacle spatial contour information of the target inspection area;

[0018] Buffer analysis is performed on the terrain elevation distribution information and the static obstacle spatial contour information to obtain the terrain safety buffer and obstacle safety buffer of the target inspection area;

[0019] Read the inspection task requirements of the target inspection area to obtain the path rule description of the target inspection area;

[0020] The terrain safety buffer, the obstacle safety buffer, and the path rule description are integrated into the navigation path measurement constraints for the target inspection area.

[0021] In a preferred embodiment, the step of performing gridded measurement on the target inspection area based on the navigation path measurement constraints to identify candidate waypoints for the target inspection area includes:

[0022] The gridded operation range of the target inspection area is defined along the three-dimensional spatial boundary contour of the target inspection area;

[0023] The meshing operation range is divided into meshes to obtain a three-dimensional orthogonal regular mesh of the meshing operation range;

[0024] Based on the positional relationship between the three-dimensional orthogonal regular grid and the target inspection area, the spatial measurement environment data is mapped to the three-dimensional orthogonal regular grid;

[0025] The feasibility of the three-dimensional orthogonal regular mesh is determined by applying the navigation path measurement constraints, and the attribute determination result of the three-dimensional orthogonal regular mesh is obtained.

[0026] Based on the attribute determination results, feasible waypoint grids of the three-dimensional orthogonal regular grid are selected;

[0027] The waypoints covered by the feasible waypoint grid are used as candidate waypoints for the target inspection area.

[0028] In a preferred embodiment, the step of analyzing the spatial relationships of the candidate waypoints based on their spatial location information to construct an initial navigation network for the target inspection area includes:

[0029] Based on the spatial location information of the candidate waypoints, spatial neighbor point tracking is performed on the candidate waypoints to confirm the adjacent candidate pairs of the candidate waypoints;

[0030] A comprehensive reachability analysis is performed on the adjacent candidate pairs to obtain their spatial relationships.

[0031] Using the candidate waypoints as network nodes and the spatial relationships as connecting edges, an initial navigation network for the target inspection area is constructed.

[0032] In a preferred embodiment, the step of performing a comprehensive reachability analysis on the adjacent candidate pairs to obtain the spatial relationship between the adjacent candidate pairs includes:

[0033] Connect the adjacent candidate pairs to form a three-dimensional straight-line segment of the adjacent candidate pairs;

[0034] Spatial visibility analysis is performed on the three-dimensional straight line segment to obtain the connectivity information of the three-dimensional straight line segment;

[0035] Based on the connectivity information, the relational attributes of the adjacent candidate pairs are assigned to obtain the spatial relationship of the adjacent candidate pairs.

[0036] In a preferred embodiment, the step of performing multi-constraint evaluation on the initial navigation network to obtain the constraint weights corresponding to the candidate waypoints, and integrating the candidate waypoints with constraint weights higher than a preset weight threshold into a globally optimal waypoint set for the target inspection area, includes:

[0037] Perform constraint cost mapping on the initial navigation network to obtain the network edge cost value of the initial navigation network;

[0038] Based on a preset dimensional importance coefficient, the edge values ​​of the network are weighted and fused to obtain the edge constraint weights of the initial navigation network. The calculation formula for the weighted fusion is as follows:

[0039] ;

[0040] In the formula, Represents the edges in the initial navigation network Edge constraint weights, Indicates the total number of constraint dimensions. Represents the edge In the The value of network edge computing in each dimension Indicates the first The dimensional importance coefficients of each dimension. This represents the preset cost sensitivity adjustment factor. This represents the preset normalization scaling factor. Represents the natural logarithm function. This represents the natural exponential function. This represents the summation operation;

[0041] Based on the edge constraint weights, constraint aggregation is performed on the network nodes of the initial navigation network to obtain the point constraint weights of the network nodes.

[0042] The point constraint weights are used as the constraint weights corresponding to the candidate waypoints;

[0043] The constraint weights are compared with a preset weight threshold, and the candidate waypoints with constraint weights higher than the preset weight threshold are integrated into the global optimal waypoint set of the target inspection area.

[0044] In a preferred embodiment, the step of performing constraint cost mapping on the initial navigation network to obtain the network edge value of the initial navigation network includes:

[0045] Spatial feature deconstruction is performed on the network edges of the initial navigation network to obtain multidimensional feature vectors of the network edges;

[0046] The multidimensional feature vector is mapped to a preset cost transformation rule to obtain the basic cost value of the multidimensional feature vector;

[0047] The network edge value of the initial navigation network is obtained by performing data reduction on the basic value.

[0048] In a preferred embodiment, the step of making adjacent waypoints in the globally optimal waypoint set traversable to obtain the navigation path for the target inspection area includes:

[0049] For adjacent waypoints in the global optimal waypoint set, straight-line segments are constructed to obtain the initial path segments of the adjacent waypoints;

[0050] Connect the initial path segments with polylines to obtain the turning path trajectory of the initial path segments;

[0051] The turning path trajectory is smoothed to obtain the navigation path for the target inspection area.

[0052] To address the aforementioned problems, the present invention also provides an automated UAV nest inspection path planning system, the system comprising:

[0053] The multi-source data fusion module is used to perform multi-source data fusion on the target inspection area to obtain the spatial measurement environment data of the target inspection area.

[0054] The constraint parsing module is used to parse the airspace constraints in the space measurement environment data and, in combination with the inspection task requirements of the target inspection area, determine the navigation path measurement constraints of the target inspection area.

[0055] The gridded waypoint identification module is used to perform gridded measurement on the target inspection area based on the navigation path measurement constraints, so as to identify candidate waypoints in the target inspection area;

[0056] The navigation network construction module is used to analyze the spatial relationships of the candidate waypoints based on their spatial location information in order to construct an initial navigation network for the target inspection area.

[0057] The multi-constraint evaluation and optimization module is used to perform multi-constraint evaluation on the initial navigation network, obtain the constraint weights corresponding to the candidate waypoints, and integrate the candidate waypoints with constraint weights higher than a preset weight threshold into the global optimal waypoint set of the target inspection area.

[0058] The waypoint connection path generation module is used to make adjacent waypoints in the global optimal waypoint set traversable to obtain the navigation path of the target inspection area.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] 1. This invention integrates multi-source data such as geographic information, remote sensing imagery, and airspace control data into the target inspection area, and completes spatial registration and structured processing to obtain complete and accurate spatial measurement environment data. Then, it analyzes the airspace constraints in the data, constructs terrain safety buffer zones and obstacle safety buffer zones in combination with the inspection task requirements, and integrates path rule descriptions to clarify the navigation path measurement constraints, providing a reliable basis for subsequent path planning and effectively improving the basic accuracy of path planning.

[0061] 2. Based on constraints, gridded measurement can accurately define the operation range and screen feasible waypoint grids, accurately identifying candidate waypoints. Combining the spatial position analysis of candidate waypoints, an initial navigation network is constructed. Then, through multi-constraint evaluation and weighted fusion calculation of edge constraint weights and aggregation to obtain point constraint weights, the globally optimal set of waypoints is screened out. Finally, adjacent waypoints are connected to ensure drivability and smoothness. The generated navigation path has both drivability and smoothness, significantly improving the operational efficiency and path reliability of UAV automated inspection. Attached Figure Description

[0062] Figure 1 This is a flowchart illustrating an automated inspection nest path planning method for unmanned aerial vehicles (UAVs) according to an embodiment of the present invention.

[0063] Figure 2 A functional block diagram of an automated drone nest inspection path planning system provided in an embodiment of the present invention;

[0064] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0065] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0066] This application provides a method for automated UAV nest inspection path planning. The execution entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0067] Reference Figure 1 The diagram shown is a flowchart illustrating an automated UAV nest inspection path planning method according to an embodiment of the present invention. In this embodiment, the automated UAV nest inspection path planning method includes:

[0068] S1. Perform multi-source data fusion on the target inspection area to obtain the spatial measurement environment data of the target inspection area;

[0069] In this embodiment of the invention, the step of fusing multi-source data of the target inspection area to obtain spatial measurement environment data of the target inspection area includes:

[0070] Geographic information data, remote sensing image data, airspace control data and inspection task parameters of the target inspection area are acquired to obtain multi-source heterogeneous data of the target inspection area;

[0071] By combining the spatial location information of the multi-source heterogeneous data, spatial registration is performed on the multi-source heterogeneous data to obtain the registered multi-source data.

[0072] The registered multi-source data is structured and integrated to obtain the spatial measurement environment data of the target inspection area.

[0073] When acquiring geographic information data for the target inspection area, high-precision topographic and geomorphological data for the area are retrieved from authoritative national or local geographic information databases. This includes topographic relief data, Cartesian coordinate data, and elevation data to ensure the data fully reflects the basic geographic features of the area. When acquiring remote sensing image data, high-resolution remote sensing satellites or dedicated aerial photography equipment with a resolution of at least 0.5 meters are used to conduct full-coverage imaging of the target inspection area. During the imaging process, flight altitude and flight path density are strictly controlled to avoid overlapping, omissions, or insufficient resolution in the images, ensuring that the images clearly show the outlines of surface objects, vegetation distribution, and the location of artificial facilities within the area. When acquiring airspace control data... Through official air traffic control information service channels or by submitting a formal application to the local air traffic control department, the scope of the inspection mission, the planned flight time, and the nature of the operation are clearly stated. This allows for the acquisition of detailed control information such as the boundary coordinates of the no-fly zone, the restricted flight altitude range, and special flight permit requirements for the area. When obtaining inspection mission parameters, the precise boundaries of the target inspection area, the key facility locations to be inspected, and the total duration of the inspection mission are determined based on specific inspection needs. All the aforementioned geographic information data, remote sensing image data, airspace control data, and inspection mission parameters are collected and their completeness is verified one by one. After confirming that no key content is missing, the data is integrated to form multi-source heterogeneous data of the target inspection area.

[0074] When performing spatial registration by combining spatial location information from multi-source heterogeneous data, the spatial location identifiers in each type of data are first extracted. For example, the latitude and longitude coordinates of terrain feature points in geographic information data, the ground coordinates corresponding to each pixel in remote sensing image data, the vertex coordinates of no-fly zone boundaries in airspace control data, and the coordinates of key inspection points in inspection task parameters. Then, the National Geodetic Coordinate System 2000 is selected as the unified reference coordinate system. This system can provide a consistent spatial reference standard for all data. Next, the coordinate transformation method is used to transform all the spatial location information in remote sensing image data, airspace control data, and inspection task parameters to the reference coordinate system. During the transformation process, the transformation parameters are repeatedly adjusted by comparing the coordinate differences of the same feature point in different data to ensure that the spatial deviation of all transformed data is controlled within a very small range, completely eliminating the spatial misalignment problem between different data. After the transformation is completed, the registered multi-source data is obtained.

[0075] When performing structured integration of registered multi-source data, the data is first categorized and organized according to data type. Geographic information data is classified as basic topographic data, including sub-items such as elevation distribution tables and landform type maps; remote sensing image data is classified as environmental visual data, including sub-items such as complete area image maps and magnified images of local details; airspace control data is classified as airspace restriction data, including sub-items such as no-fly zone lists and restricted flight altitude tables; and inspection task parameters are classified as task requirement data, including sub-items such as inspection range tables and lists of key inspection objects. Then, spatial relationships are established between these data types. Elevation information in the basic topographic data is linked using the same latitude and longitude coordinates. The system associates the corresponding image regions in the visual data of the environment with the information, enabling the direct acquisition of the elevation value of a specific image location when viewing it. Simultaneously, it associates the airspace restriction data with the task requirement data through the coordinates of the inspection range, clarifying the airspace restriction requirements within the inspection range. Finally, according to a preset structured format, the key information of each type of data is filled into the corresponding fields one by one. For example, the specific elevation value and landform description are filled in the "Specific Content of Data" field of the terrain basic data, and the coordinate range and restricted flight altitude of the no-fly zone are filled in the "Specific Content of Data" field of the airspace restriction data. After completing the organization and storage of all the data, the spatial measurement environment data of the target inspection area is obtained.

[0076] The beneficial effects are as follows: by acquiring multi-source heterogeneous data through step-by-step and refined operations, the data ensures that it covers all dimensions of information such as geography, imagery, airspace control, and mission requirements of the target inspection area without any omissions; by eliminating spatial deviations between different data through strict spatial registration, the data is ensured to maintain consistency within the same spatial framework; and by integrating the system in a structured manner, the data is clearly classified, closely related, and formatted in a standardized manner. The resulting spatial measurement environment data is complete, spatially accurate, and structurally clear, providing high-quality data support for subsequent analysis of airspace constraints and determination of navigation path measurement constraints, thus ensuring the accuracy and reliability of UAV automated inspection nest path planning from the source.

[0077] S2. Analyze the airspace constraints in the space measurement environment data and combine them with the inspection task requirements of the target inspection area to determine the navigation path measurement constraints of the target inspection area.

[0078] In this embodiment of the invention, the step of parsing the airspace constraints in the spatial measurement environment data and combining them with the inspection task requirements of the target inspection area to determine the navigation path measurement constraints of the target inspection area includes:

[0079] Spatial feature deconstruction is performed on the spatial measurement environment data to obtain the terrain elevation distribution information and static obstacle spatial contour information of the target inspection area;

[0080] Buffer analysis is performed on the terrain elevation distribution information and the static obstacle spatial contour information to obtain the terrain safety buffer and obstacle safety buffer of the target inspection area;

[0081] Read the inspection task requirements of the target inspection area to obtain the path rule description of the target inspection area;

[0082] The terrain safety buffer, the obstacle safety buffer, and the path rule description are integrated into the navigation path measurement constraints for the target inspection area.

[0083] When performing spatial feature deconstruction on the spatial measurement environment data, the elevation values ​​corresponding to each coordinate point are extracted from the terrain basic data in the spatial measurement environment data. The elevation values ​​are statistically analyzed by latitude and longitude partition, and the highest, lowest, and average elevations in each partition are recorded. At the same time, the coordinate intervals where the elevation values ​​change drastically are marked, forming terrain elevation distribution information that can clearly reflect the elevation conditions of different locations in the target inspection area. From the environmental visual data in the spatial measurement environment data, the stationary artificial facilities and natural obstacles in the images are identified by analyzing remote sensing images frame by frame. The outer contour coordinates of these static obstacles are delineated, and the contour boundary range of each obstacle is determined, forming the spatial contour information of static obstacles in the target inspection area.

[0084] When performing buffer analysis on the terrain elevation distribution information and the static obstacle spatial contour information, for areas in the terrain elevation distribution information where the elevation change exceeds a safety threshold, a fixed safety distance is extended outward from the boundary coordinates of that area to form a closed polygonal region. This region is the terrain safety buffer zone of the target inspection area, and the boundary coordinates of the buffer zone must strictly correspond to the boundary of the dangerous area in the terrain elevation distribution information. For each obstacle in the static obstacle spatial contour information, a fixed safety distance is extended outward from the coordinates of the outermost contour point of the obstacle to form a closed polygonal region surrounding the obstacle. This region is the obstacle safety buffer zone of the target inspection area. Each obstacle corresponds to a unique obstacle safety buffer zone and is marked with a clear coordinate range.

[0085] When reading the inspection task requirements of the target inspection area, specific requirements related to the inspection path are extracted from the task requirement data in the space measurement environment data. These include the direction that the inspection route must follow, the key inspection points that must be covered, the specific areas that the flight path must avoid, and the length limit that the overall path must meet. These specific requirements are then organized into clear and well-defined text to obtain the path rule description of the target inspection area.

[0086] When integrating the terrain safety buffer, the obstacle safety buffer, and the path rule description into the navigation path measurement constraints for the target inspection area, the coordinate ranges of the terrain safety buffer and the obstacle safety buffer are first presented in tabular form, clearly indicating the name, boundary coordinates, and corresponding safety restrictions of each buffer. Then, the path rule descriptions are appended to the table as entries, ensuring that each rule corresponds to a specific path planning requirement. Finally, the table and entry contents are checked to ensure that the spatial constraints of the terrain safety buffer and the obstacle safety buffer do not conflict with the task requirements of the path rule description, forming a complete document containing both spatial and task constraints. This document constitutes the navigation path measurement constraints for the target inspection area.

[0087] The beneficial effects are: accurately obtaining terrain elevation distribution information and static obstacle spatial contour information through spatial feature deconstruction, providing precise data support for the subsequent establishment of safety buffer zones; clarifying the safe avoidance range of terrain and obstacles through buffer zone analysis, eliminating spatial safety hazards in path planning; obtaining path rule descriptions that fit actual operations by reading inspection task requirements; and finally integrating the navigation path measurement constraints to cover both the safety limitations of the spatial environment and the actual requirements of the task, ensuring that the constraints are comprehensive and accurate, providing a clear basis for subsequent grid-based measurement and waypoint selection based on the constraints, and ensuring that subsequent path planning meets safety standards and operational requirements.

[0088] S3. Based on the navigation path measurement constraints, perform gridded measurement on the target inspection area to identify candidate waypoints for the target inspection area;

[0089] In this embodiment of the invention, the step of performing gridded measurement on the target inspection area based on the navigation path measurement constraints to identify candidate waypoints for the target inspection area includes:

[0090] The gridded operation range of the target inspection area is defined along the three-dimensional spatial boundary contour of the target inspection area;

[0091] The meshing operation range is divided into meshes to obtain a three-dimensional orthogonal regular mesh of the meshing operation range;

[0092] Based on the positional relationship between the three-dimensional orthogonal regular grid and the target inspection area, the spatial measurement environment data is mapped to the three-dimensional orthogonal regular grid;

[0093] The feasibility of the three-dimensional orthogonal regular mesh is determined by applying the navigation path measurement constraints, and the attribute determination result of the three-dimensional orthogonal regular mesh is obtained.

[0094] Based on the attribute determination results, feasible waypoint grids of the three-dimensional orthogonal regular grid are selected;

[0095] The waypoints covered by the feasible waypoint grid are used as candidate waypoints for the target inspection area.

[0096] When defining the gridded operation range of the target inspection area along the three-dimensional spatial boundary contour of the target inspection area, the planar boundary coordinates and elevation boundary values ​​of the target inspection area are first extracted from the spatial measurement environment data. These boundary information are then transformed into a visualized three-dimensional spatial boundary contour using a three-dimensional modeling tool. This contour completely encloses all the spaces to be inspected in the target inspection area. Then, based on the three-dimensional spatial boundary contour, a very small fixed distance is extended outward to form a closed three-dimensional cube space. This three-dimensional cube space is the gridded operation range of the target inspection area, and the longitude start and end values, latitude start and end values, and elevation start and end values ​​of this range need to be recorded.

[0097] When dividing the meshed operation range into a three-dimensional orthogonal regular mesh, the three-dimensional orthogonal direction is first determined, and a fixed division interval is set in each orthogonal direction. Then, the meshed operation range is divided into several slices parallel to the latitude-elevation plane along the longitude direction according to the division interval, several slices parallel to the longitude-elevation plane along the latitude direction according to the division interval, and several slices parallel to the longitude-latitude plane along the elevation direction according to the division interval. These slices intersect each other, dividing the meshed operation range into multiple three-dimensional cube units of uniform size and the same shape. Each cube unit is a three-dimensional orthogonal regular mesh, and a unique coordinate code is assigned to each mesh to distinguish different meshes.

[0098] When mapping the spatial measurement environment data to the three-dimensional orthogonal regular grid based on the positional relationship between the three-dimensional orthogonal regular grid and the target inspection area, the positional relationship is first established—by comparing the coordinate code of each three-dimensional orthogonal regular grid with the spatial coordinate range of each piece of information in the spatial measurement environment data, the environmental data corresponding to each grid is determined; then, the elevation values ​​in the terrain elevation data that belong to the coordinate range of the grid, the obstacle fragment information in the static obstacle spatial contour data that belong to the coordinate range of the grid, and the control requirements in the airspace control data that belong to the coordinate range of the grid are filled into the information storage unit of the grid one by one, so that each three-dimensional orthogonal regular grid is accompanied by complete environmental data corresponding to its spatial position, thus completing the mapping of spatial measurement environment data to the three-dimensional orthogonal regular grid.

[0099] When applying the navigation path measurement constraints to determine the feasibility of the three-dimensional orthogonal regular grid and obtaining the attribute determination result of the three-dimensional orthogonal regular grid, the environmental data attached to each three-dimensional orthogonal regular grid is retrieved one by one to check whether the grid is within the coordinate range of the terrain safety buffer or obstacle safety buffer. Then, the path rule description is compared to check whether the grid meets the task constraints such as path direction and coverage requirements. If the grid is neither within any safety buffer nor meets all path rule requirements, the grid is determined to be "feasible". If the grid is within a safety buffer or does not meet any path rule requirements, the grid is determined to be "infeasible". The "feasible" or "infeasible" determination result of each grid is recorded in correspondence with the grid coordinate code to form the attribute determination result of the three-dimensional orthogonal regular grid.

[0100] When selecting feasible waypoint grids for the 3D orthogonal regular grid based on the attribute determination results, the attribute determination results of all 3D orthogonal regular grids are first traversed, and grids with the determination result of "feasible" are extracted. Then, these "feasible" grids are checked a second time to confirm the matching of their attached environmental data with the navigation path measurement constraints, and grids that do not meet the constraints are removed. Finally, the remaining "feasible" grids that meet the constraints are compiled into a list, which includes the coordinate code, spatial coordinate range and attached environmental data of each grid. These grids are the feasible waypoint grids for the 3D orthogonal regular grid.

[0101] When using the waypoints covered by the feasible waypoint grid as candidate waypoints for the target inspection area, first determine the center coordinates of each feasible waypoint grid, which are the waypoints covered by that feasible waypoint grid; then extract the center coordinates of all feasible waypoint grids one by one, and record the three-dimensional coordinate values ​​of each waypoint and the corresponding feasible waypoint grid code; finally, organize these waypoints into a set according to coordinate order, ensuring that there are no duplicate waypoints in the set, and this set is the candidate waypoint for the target inspection area.

[0102] The beneficial effects are as follows: by accurately defining the scope of the gridded operation, it ensures that the grid segmentation does not miss any areas to be inspected and does not include irrelevant spaces; by using standardized three-dimensional orthogonal grid segmentation, the grid has a unified standard and facilitates subsequent data processing; by accurately mapping environmental data, it provides a comprehensive basis for grid feasibility determination; by strictly judging and screening constraints, it ensures that feasible waypoint grids meet safety and mission requirements; and the final candidate waypoints accurately correspond to spatial locations that meet the constraints, providing a high-quality node foundation for the subsequent construction of the initial navigation network, and ensuring the safety and practicality of the navigation network.

[0103] S4. Based on the spatial location information of the candidate waypoints, analyze the spatial relationships of the candidate waypoints to construct an initial navigation network for the target inspection area;

[0104] In this embodiment of the invention, the step of analyzing the spatial relationships of the candidate waypoints based on their spatial location information to construct an initial navigation network for the target inspection area includes:

[0105] Based on the spatial location information of the candidate waypoints, spatial neighbor point tracking is performed on the candidate waypoints to confirm the adjacent candidate pairs of the candidate waypoints;

[0106] A comprehensive reachability analysis is performed on the adjacent candidate pairs to obtain their spatial relationships.

[0107] Using the candidate waypoints as network nodes and the spatial relationships as connecting edges, an initial navigation network for the target inspection area is constructed.

[0108] The reachability comprehensive analysis of the adjacent candidate pairs to obtain the spatial relationship of the adjacent candidate pairs includes:

[0109] Connect the adjacent candidate pairs to form a three-dimensional straight-line segment of the adjacent candidate pairs;

[0110] Spatial visibility analysis is performed on the three-dimensional straight line segment to obtain the connectivity information of the three-dimensional straight line segment;

[0111] Based on the connectivity information, the relational attributes of the adjacent candidate pairs are assigned to obtain the spatial relationship of the adjacent candidate pairs.

[0112] When performing spatial proximity tracking on the candidate waypoints to confirm adjacent candidate pairs, the three-dimensional coordinates of each candidate waypoint are first extracted from the candidate waypoint set, including longitude, latitude, and elevation information. A fixed proximity distance threshold is set based on the UAV's flight accuracy requirements and the spatial scale of the target inspection area. This threshold must ensure that the UAV can safely fly from one waypoint to a neighboring waypoint without significant route adjustments. Then, for each candidate waypoint, the three-dimensional straight-line distance between it and all other candidate waypoints is calculated. The calculation of the straight-line distance between two points using spatial coordinates ensures that the calculation results accurately reflect the actual spatial distance. The calculated distance is compared with the set proximity distance threshold. If the distance between two candidate waypoints is less than or equal to the threshold, these two candidate waypoints are identified as an adjacent candidate pair. During the confirmation process, it is necessary to avoid repeatedly recording the same adjacent candidate pair. For example, if waypoint A and waypoint B have already been recorded as an adjacent candidate pair, waypoint B and waypoint A will not be recorded again. Finally, a list containing all adjacent candidate pairs is compiled, clearly marking the two candidate waypoints and their corresponding three-dimensional coordinates for each adjacent candidate pair.

[0113] When connecting adjacent candidate pairs to form a three-dimensional straight line segment for each pair, the three-dimensional coordinates of the two candidate waypoints are extracted for each pair. For example, the longitude of waypoint P is P1, the latitude is P2, and the elevation is P3, while the longitude of waypoint Q is Q1, the latitude is Q2, and the elevation is Q3. Using the three-dimensional coordinates of waypoint P as the starting point and the three-dimensional coordinates of waypoint Q as the ending point, a straight line is used to connect these two coordinate points in three-dimensional space. This straight line connecting the starting point and the ending point is the three-dimensional straight line segment corresponding to the adjacent candidate pair. To facilitate subsequent management and analysis, a unique identifier number needs to be assigned to each three-dimensional straight line segment. This number is consistent with the identifier of the corresponding adjacent candidate pair. At the same time, the starting point coordinates, ending point coordinates, and straight line length of each three-dimensional straight line segment are recorded. The length is calculated from the starting point and ending point coordinates to ensure that the information of each three-dimensional straight line segment is complete and traceable.

[0114] When performing spatial visibility analysis on the three-dimensional straight flight segment to obtain its connectivity information, the terrain elevation distribution information and static obstacle spatial contour information of the area traversed by the three-dimensional straight flight segment are first retrieved from the spatial measurement environment data. Multiple sampling points are selected at uniform intervals along the extension direction of the three-dimensional straight flight segment. The interval size must ensure comprehensive coverage of the flight segment without omitting key areas. For example, a sufficient number of sampling points are selected based on the flight segment length so that the distance between adjacent sampling points accurately reflects the environmental characteristics of the area traversed by the flight segment. For each sampling point, its three-dimensional coordinates are extracted, and the actual terrain elevation of the sampling point is checked against the terrain elevation distribution information to see if it is lower than the three-dimensional coordinates. If the actual terrain elevation at the sampling point is higher than the elevation of the flight segment, it indicates that terrain occlusion exists. At the same time, the sampling point is checked against the spatial contour information of static obstacles to see if it is within the spatial contour range of any static obstacle. If it is, it indicates that obstacle occlusion exists. If all sampling points are free from terrain occlusion and obstacle occlusion, the connectivity information of the three-dimensional straight flight segment is determined to be connectable. If any sampling point is obstructed by terrain or obstacle, the connectivity information of the three-dimensional straight flight segment is determined to be disconnected. The identifier number of each three-dimensional straight flight segment and the corresponding connectable or disconnected result are recorded to form a connectivity information list of the three-dimensional straight flight segment.

[0115] When assigning relational attributes to adjacent candidate pairs based on the connectivity information to obtain the spatial relationship of the adjacent candidate pairs, the connectivity information of the three-dimensional straight-line segments is first associated with the corresponding adjacent candidate pairs. This association is achieved through the consistency between the segment identifier number and the identifier of the adjacent candidate pair. If the connectivity information of the three-dimensional straight-line segments corresponding to an adjacent candidate pair is connectable, then a passable connection relational attribute is assigned to the adjacent candidate pair, indicating that the two candidate waypoints can pass safely through the straight-line segments. If the connectivity information of the three-dimensional straight-line segments corresponding to an adjacent candidate pair is not connectable, then a non-passable connection relational attribute is assigned to the adjacent candidate pair, indicating that there is an obstruction between the two candidate waypoints and they cannot pass safely through the straight-line segments. Subsequently, the two waypoint identifiers and corresponding relational attributes of each pair of adjacent candidate pairs are integrated together to form the spatial relationship of the adjacent candidate pair. Each adjacent candidate pair corresponds to a unique spatial relationship record, and the record content must clearly reflect the feasibility of passage between waypoints.

[0116] When constructing the initial navigation network for the target inspection area using the candidate waypoints as network nodes and the spatial relationships as connecting edges, all candidate waypoints are first treated as independent network nodes, each corresponding to a candidate waypoint. The node stores the core information of the candidate waypoint, such as its 3D coordinates and identifier number. Then, spatial relationships with traversable connection attributes are selected from all spatial relationships and used as connecting edges. Each connecting edge connects two network nodes, namely the network nodes corresponding to the two candidate waypoints of the adjacent candidate pair to which the spatial relationship belongs. The corresponding spatial relationship identifier and traversable connection attribute are recorded in the connecting edge. Subsequently, network modeling is used to combine all network nodes and the selected connecting edges according to their corresponding relationships to form a structured network model. In this model, each network node can be associated with other nodes that meet the traversable connection conditions through connecting edges, and there are no invalid or erroneous connections. This network model, which includes network nodes, traversable connecting edges, and corresponding association relationships, is the initial navigation network for the target inspection area.

[0117] The beneficial effects are as follows: spatial proximity tracking can accurately screen adjacent candidate pairs between candidate waypoints, ensuring that the adjacency relationship meets the actual flight requirements of UAVs; through the construction of three-dimensional straight-line segments and spatial visibility analysis, the feasibility of passage between waypoints can be accurately determined, avoiding the risk of occlusion in subsequent path planning; by assigning relational attributes, the spatial relationship between waypoints is clarified, providing a clear basis for the selection of connecting edges; and the final initial navigation network has clear nodes and feasible connecting edges, which can fully reflect the effective passage relationship between candidate waypoints, providing a reliable network foundation for subsequent multi-constraint evaluation and path optimization, and ensuring the safety and efficiency of UAV inspection path planning.

[0118] S5. Perform multi-constraint evaluation on the initial navigation network to obtain the constraint weights corresponding to the candidate waypoints, and integrate the candidate waypoints with constraint weights higher than a preset weight threshold into the global optimal waypoint set of the target inspection area.

[0119] In this embodiment of the invention, the step of performing multi-constraint evaluation on the initial navigation network to obtain the constraint weights corresponding to the candidate waypoints, and integrating the candidate waypoints with constraint weights higher than a preset weight threshold into a globally optimal waypoint set for the target inspection area, includes:

[0120] Perform constraint cost mapping on the initial navigation network to obtain the network edge cost value of the initial navigation network;

[0121] Based on a preset dimensional importance coefficient, the edge values ​​of the network are weighted and fused to obtain the edge constraint weights of the initial navigation network. The calculation formula for the weighted fusion is as follows:

[0122] ;

[0123] In the formula, Represents the edges in the initial navigation network Edge constraint weights, Indicates the total number of constraint dimensions. Represents the edge In the The value of network edge computing in each dimension Indicates the first The dimensional importance coefficients of each dimension. This represents the preset cost sensitivity adjustment factor. This represents the preset normalization scaling factor. Represents the natural logarithm function. This represents the natural exponential function. This represents the summation operation;

[0124] Based on the edge constraint weights, constraint aggregation is performed on the network nodes of the initial navigation network to obtain the point constraint weights of the network nodes.

[0125] The point constraint weights are used as the constraint weights corresponding to the candidate waypoints;

[0126] The constraint weights are compared with a preset weight threshold, and the candidate waypoints with constraint weights higher than the preset weight threshold are integrated into the global optimal waypoint set of the target inspection area.

[0127] The step of performing constraint cost mapping on the initial navigation network to obtain the network edge value of the initial navigation network includes:

[0128] Spatial feature deconstruction is performed on the network edges of the initial navigation network to obtain multidimensional feature vectors of the network edges;

[0129] The multidimensional feature vector is mapped to a preset cost transformation rule to obtain the basic cost value of the multidimensional feature vector;

[0130] The basic value is reduced by data to obtain the network edge value of the initial navigation network.

[0131] When performing constraint cost mapping on the initial navigation network to obtain the network edge value, the network edges of the initial navigation network are first deconstructed in terms of spatial features. Each network edge is an edge connecting two candidate waypoints. The three-dimensional spatial coordinate range of each edge is extracted, that is, the longitude, latitude and elevation information of the starting and ending candidate waypoints of the edge. The actual length of the edge is extracted and the straight-line distance between the two points is calculated by the three-dimensional coordinates of the starting and ending points. The environmental features of the area traversed by the edge are extracted, including whether the edge extension path is close to the terrain safety buffer, whether it is close to the obstacle safety buffer, and whether the edge extension direction is consistent with the path rule direction in the inspection task requirements. These extracted three-dimensional spatial coordinate range, actual length and environmental feature information are organized into a set of structured data in a fixed order. This set of structured data is the multi-dimensional feature vector of the network edge.

[0132] When mapping a multidimensional feature vector to a preset cost transformation rule to obtain the base cost value of the multidimensional feature vector, the preset cost transformation rule needs to be formulated in advance. The rule content includes: if the actual length of an edge exceeds the preset safe length range, each additional unit of length beyond the limit corresponds to a fixed cost value increment; if the extension path of an edge is within a preset distance outside the terrain safety buffer zone, it corresponds to a fixed cost value; if the extension path of an edge is within a preset distance outside the obstacle safety buffer zone, it corresponds to a fixed cost value higher than that of the terrain safety buffer zone; and if the deviation angle between the extension direction of the edge and the direction required by the path rule in the inspection task requirements increases by a fixed angle, it corresponds to a fixed cost value increment. Each feature item in the multidimensional feature vector is matched with the corresponding clause in the cost transformation rule to find the cost value corresponding to each feature item. The cost values ​​corresponding to all feature items are accumulated, and the accumulated result is the base cost value of the multidimensional feature vector.

[0133] When reducing the base cost value to obtain the network edge cost value of the initial navigation network, the base cost value of all network edges in the initial navigation network is first collected, and the maximum and minimum values ​​are recorded. The difference between the maximum and minimum values ​​is calculated. If the difference is not zero, the base cost value of each network edge is processed by subtracting the minimum value from the base cost value and then dividing by the difference between the maximum and minimum values ​​to obtain the normalized cost value. If the difference is zero, it means that the base cost value of all network edges is exactly the same, and the base cost value is directly used as the normalized cost value. This normalization process ensures that the cost value of all network edges is within the same numerical range. The normalized cost value is the network edge cost value of the initial navigation network.

[0134] When obtaining the edge constraint weights of the initial navigation network by weighted fusion of network edge values ​​based on preset dimensional importance coefficients, the total number of constraint dimensions to be included in the evaluation is first determined. These dimensions include terrain safety compliance, obstacle safety compliance, path direction matching degree, length compliance, etc. Each constraint dimension corresponds to a preset dimensional importance coefficient, the magnitude of which is set according to the degree of influence of that dimension on UAV flight safety and inspection task completion effect; the greater the influence, the higher the coefficient value. At the same time, a preset cost sensitivity adjustment factor is determined, which is used to control the degree of influence of changes in network edge values ​​on the final result. Also, a preset normalization scaling factor is determined. The factor is set based on the maximum value of the natural logarithm result in subsequent calculations under all possible conditions to ensure that the final result falls within the preset range. For each network edge, the network edge value in each constraint dimension is first multiplied by the cost sensitivity adjustment factor, and then the natural exponential function value of the product is calculated to obtain the exponential result in that dimension. The exponential result of each dimension is multiplied by the dimension importance coefficient of the corresponding dimension, and the product results of all dimensions are added together to obtain the sum. The natural logarithm result of this sum is calculated by adding one, and finally the natural logarithm result is divided by the normalization scaling factor. The obtained value is the edge constraint weight of that network edge in the initial navigation network.

[0135] When performing constraint aggregation on network nodes of the initial navigation network based on edge constraint weights to obtain the point constraint weights of the network nodes, each network node corresponds to a candidate waypoint. First, find all network edges directly connected to the network node, that is, all network edges with the network node as the starting point or ending point. Collect the edge constraint weights corresponding to these connected network edges. If there is only one connected network edge, the edge constraint weight of this edge is directly used as the point constraint weight of the network node. If there are two or more connected network edges, the edge constraint weights of all connected edges are added together to obtain the total weight. Then, the total weight is divided by the total number of connected edges to obtain the average weight. This average weight is the point constraint weight of the network node.

[0136] When using point constraint weights as constraint weights for candidate waypoints, since each network node in the initial navigation network corresponds one-to-one with each candidate waypoint in the target inspection area (i.e., one network node uniquely corresponds to one candidate waypoint), the point constraint weights of each network node are directly assigned to their corresponding candidate waypoints. After the assignment, each candidate waypoint has a corresponding value, which is the constraint weight corresponding to the candidate waypoint.

[0137] When comparing constraint weights with preset weight thresholds and integrating candidate waypoints with constraint weights higher than the preset weight thresholds into a globally optimal waypoint set for the target inspection area, a preset weight threshold is first set according to the safety standards and mission completion accuracy requirements of UAV automated inspection. The threshold value must ensure that only candidate waypoints with sufficiently high constraint weights can meet the flight safety requirements and inspection mission requirements. The constraint weights corresponding to each candidate waypoint are extracted one by one, and the constraint weights are compared with the preset weight thresholds. If the constraint weight of a candidate waypoint is higher than the preset weight threshold, the candidate waypoint is selected. All selected candidate waypoints are collected and arranged according to their spatial position in the target inspection area to form an ordered waypoint set, which is the globally optimal waypoint set for the target inspection area.

[0138] The beneficial effects are as follows: by deconstructing the spatial features of network edges, key information of edges can be fully extracted, ensuring that multi-dimensional feature vectors completely reflect the attributes of edges. Combined with preset cost transformation rules, features are quantified into basic cost values ​​and unified within a range through data reduction, making the cost values ​​of network edges accurate and comparable. In the weighted fusion process, the impact of key constraints on safety and mission is highlighted by setting dimensional importance coefficients. The degree of influence of cost value changes is precisely controlled by cost sensitivity adjustment factors. The normalization scaling factor is used to keep the edge constraint weights within a reasonable range, ensuring that the edge constraint weights can comprehensively reflect the compliance of edges. Then, through constraint aggregation, the edge weights are transformed into node constraint weights, so that the constraint weights of candidate waypoints can reflect the overall quality of their connecting edges. Finally, candidate waypoints that meet the requirements are selected through preset thresholds to form a globally optimal waypoint set. Each step of the entire process has clear operation and quantitative basis, effectively ensuring the accuracy and reliability of multi-constraint evaluation, providing a high-quality waypoint foundation for subsequent navigation path generation, and ensuring that the UAV inspection path is safe and meets mission requirements.

[0139] S6. Connect adjacent waypoints in the global optimal waypoint set to obtain the navigation path of the target inspection area.

[0140] In this embodiment of the invention, the step of making adjacent waypoints in the globally optimal waypoint set traversable to obtain the navigation path for the target inspection area includes:

[0141] For adjacent waypoints in the global optimal waypoint set, straight-line segments are constructed to obtain the initial path segments of the adjacent waypoints;

[0142] Connect the initial path segments with polylines to obtain the turning path trajectory of the initial path segments;

[0143] The turning path trajectory is smoothed to obtain the navigation path for the target inspection area.

[0144] When constructing the initial path segment of adjacent waypoints in the globally optimal waypoint set, the three-dimensional coordinates of all waypoints, including longitude, latitude, and elevation information, are first extracted from the globally optimal waypoint set. Based on the path direction specified in the inspection task requirements, such as along the extension direction of a specific facility or a preset order from the starting point to the end point, all waypoints are spatially sorted to determine the preceding and following adjacent waypoints for each waypoint. For each group of adjacent waypoints after sorting, such as waypoint M and waypoint N, a straight line is drawn in three-dimensional space, starting from the three-dimensional coordinates of the preceding waypoint (waypoint M) and ending at the three-dimensional coordinates of the following waypoint (waypoint N). This straight line is the straight path segment corresponding to that group of adjacent waypoints. A unique identifier is assigned to each straight path segment, recording its starting coordinates, ending coordinates, and straight length. All straight path segments are arranged according to the waypoint sorting order, with each straight path segment serving as an independent path unit, collectively forming the initial path segment of adjacent waypoints.

[0145] When obtaining the turning trajectory of the initial path segment by connecting the initial path segments with polylines, the connection relationship between two adjacent initial path segments is first analyzed to find the endpoint of the previous initial path segment, which is also the starting point of the next initial path segment, and is denoted as the connection endpoint. Based on the minimum turning radius of the UAV, which is determined according to the UAV's performance parameters, ensuring that the turning process does not exceed the limitations of the fuselage structure and flight stability requirements, the three-dimensional coordinates of the turning point are determined between the extension lines of the previous and next initial path segments at the connection endpoint—the previous flight segment at the connection endpoint is calculated. Using the tangent direction of the first initial path segment and the tangent direction of the next segment as a reference, along the angle bisector of the two tangent directions, determine the position of the turning point according to the length of the minimum turning radius; connect the end point of the previous initial path segment with the turning point with a straight line, and then connect the turning point with the starting point of the next initial path segment with a straight line to form a broken line containing the turning process; connect all adjacent initial path segments in this way to form a continuous path whole. This continuous path whole containing the turning broken line is the turning path trajectory of the initial path segment, and at the same time record the coordinates and turning angle of each turning point.

[0146] When smoothing the turning path trajectory to obtain the navigation path for the target inspection area, a Bézier curve fitting method is used to smooth and optimize the turning path trajectory. First, key control points are selected from the turning path trajectory—the starting and ending points of all initial path segments and all turning points in the turning path trajectory are used as control points, arranged in the order of trajectory extension. Based on the control point sequence, the Bézier curve fitting rules are applied to calculate the curve equation parameters: Three adjacent control points, such as control points A, B, and C, are grouped together, with control point A as the curve's starting point, control point C as the curve's ending point, and control point B as the curve shape adjustment point. The curve is adjusted by adjusting control point B... The degree of traction on the curve is such that when the curve passes through control points A and C, the tangent direction is consistent with the tangent direction of the original turning path trajectory, and the overall curvature of the curve does not exceed the maximum allowable curvature of the UAV flight, thus avoiding overload during UAV flight. Curve fitting is performed sequentially on all control point sequences, and the fitted multiple segments of Bézier curves are connected to form a continuous and smooth curve. Finally, it is checked whether the smoothed curve maintains a safe distance from the terrain safety buffer and obstacle safety buffer. If there is insufficient distance in some areas, the control point positions are finely adjusted and refitted until the curve fully meets the navigation path measurement constraints. This final smooth curve is the navigation path for the target inspection area.

[0147] The beneficial effects are as follows: constructing straight flight segments enables the rapid determination of basic paths between adjacent waypoints, ensuring precise correspondence between initial path segments and waypoint spatial positions; connecting independent initial path segments through polyline paths solves the connection problem, and the turning design complies with the flight performance limitations of UAVs, avoiding safety risks during turning; path smoothing eliminates the abruptness of polyline turning, making the curvature change of the final navigation path gradual, reducing the difficulty of UAV control and energy consumption during flight; the final navigation path not only meets the connection requirements of the globally optimal waypoints but also meets the flight safety and stability requirements of UAVs, and can be directly used for automated UAV inspection operations, improving the smoothness and reliability of the inspection process.

[0148] like Figure 2 The diagram shown is a functional block diagram of an automated inspection nest path planning system for unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention.

[0149] The UAV automated inspection nest path planning system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the UAV automated inspection nest path planning system 100 may include a multi-source data fusion module 101, a constraint condition parsing module 102, a gridded waypoint identification module 103, a navigation network construction module 104, a multi-constraint evaluation and optimization module 105, and a waypoint connection path generation module 106. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0150] In this embodiment, the functions of each module / unit are as follows:

[0151] The multi-source data fusion module 101 is used to perform multi-source data fusion on the target inspection area to obtain the spatial measurement environment data of the target inspection area.

[0152] The constraint parsing module 102 is used to parse the airspace constraints in the space measurement environment data and, in combination with the inspection task requirements of the target inspection area, determine the navigation path measurement constraints of the target inspection area.

[0153] The gridded waypoint identification module 103 is used to perform gridded measurement on the target inspection area based on the navigation path measurement constraints, so as to identify candidate waypoints in the target inspection area.

[0154] The navigation network construction module 104 is used to analyze the spatial relationship of the candidate waypoints based on their spatial location information in order to construct an initial navigation network for the target inspection area.

[0155] The multi-constraint evaluation and optimization module 105 is used to perform multi-constraint evaluation on the initial navigation network, obtain the constraint weights corresponding to the candidate waypoints, and integrate the candidate waypoints with constraint weights higher than a preset weight threshold into the global optimal waypoint set of the target inspection area.

[0156] The waypoint connection path generation module 106 is used to make adjacent waypoints in the global optimal waypoint set traversable to obtain the navigation path of the target inspection area.

[0157] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0158] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0159] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0160] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0161] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for automated inspection nest path planning by unmanned aerial vehicles (UAVs), characterized in that, The method includes: S1. Perform multi-source data fusion on the target inspection area to obtain the spatial measurement environment data of the target inspection area; S2. Analyze the airspace constraints in the space measurement environment data and combine them with the inspection task requirements of the target inspection area to determine the navigation path measurement constraints of the target inspection area. S3. Based on the navigation path measurement constraints, perform gridded measurement on the target inspection area to identify candidate waypoints for the target inspection area; S4. Based on the spatial location information of the candidate waypoints, analyze the spatial relationships of the candidate waypoints to construct an initial navigation network for the target inspection area; S5. Perform multi-constraint evaluation on the initial navigation network to obtain the constraint weights corresponding to the candidate waypoints, and integrate the candidate waypoints with constraint weights higher than a preset weight threshold into the global optimal waypoint set of the target inspection area. S6. Connect adjacent waypoints in the global optimal waypoint set to obtain the navigation path of the target inspection area.

2. The method for automated inspection nest path planning of unmanned aerial vehicles (UAVs) as described in claim 1, characterized in that, The process of fusing multi-source data into the target inspection area to obtain spatial measurement environment data for the target inspection area includes: Geographic information data, remote sensing image data, airspace control data and inspection task parameters of the target inspection area are acquired to obtain multi-source heterogeneous data of the target inspection area; By combining the spatial location information of the multi-source heterogeneous data, spatial registration is performed on the multi-source heterogeneous data to obtain the registered multi-source data. The registered multi-source data is structured and integrated to obtain the spatial measurement environment data of the target inspection area.

3. The method for automated inspection nest path planning of unmanned aerial vehicles (UAVs) as described in claim 1, characterized in that, The process of analyzing the airspace constraints in the spatial measurement environment data and combining them with the inspection task requirements of the target inspection area to determine the navigation path measurement constraints of the target inspection area includes: Spatial feature deconstruction is performed on the spatial measurement environment data to obtain the terrain elevation distribution information and static obstacle spatial contour information of the target inspection area; Buffer analysis is performed on the terrain elevation distribution information and the static obstacle spatial contour information to obtain the terrain safety buffer and obstacle safety buffer of the target inspection area; Read the inspection task requirements of the target inspection area to obtain the path rule description of the target inspection area; The terrain safety buffer, the obstacle safety buffer, and the path rule description are integrated into the navigation path measurement constraints for the target inspection area.

4. The method for automated inspection nest path planning of unmanned aerial vehicles (UAVs) as described in claim 1, characterized in that, The step of performing gridded measurement on the target inspection area based on the navigation path measurement constraints to identify candidate waypoints for the target inspection area includes: The gridded operation range of the target inspection area is defined along the three-dimensional spatial boundary contour of the target inspection area; The meshing operation range is divided into meshes to obtain a three-dimensional orthogonal regular mesh of the meshing operation range; Based on the positional relationship between the three-dimensional orthogonal regular grid and the target inspection area, the spatial measurement environment data is mapped to the three-dimensional orthogonal regular grid; The feasibility of the three-dimensional orthogonal regular mesh is determined by applying the navigation path measurement constraints, and the attribute determination result of the three-dimensional orthogonal regular mesh is obtained. Based on the attribute determination results, feasible waypoint grids of the three-dimensional orthogonal regular grid are selected; The waypoints covered by the feasible waypoint grid are used as candidate waypoints for the target inspection area.

5. The method for automated inspection nest path planning of unmanned aerial vehicles (UAVs) as described in claim 1, characterized in that, The step of analyzing the spatial relationships of the candidate waypoints based on their spatial location information to construct an initial navigation network for the target inspection area includes: Based on the spatial location information of the candidate waypoints, spatial neighbor point tracking is performed on the candidate waypoints to confirm the adjacent candidate pairs of the candidate waypoints; A comprehensive reachability analysis is performed on the adjacent candidate pairs to obtain their spatial relationships. Using the candidate waypoints as network nodes and the spatial relationships as connecting edges, an initial navigation network for the target inspection area is constructed.

6. The method for automated inspection nest path planning of unmanned aerial vehicles (UAVs) as described in claim 5, characterized in that, The reachability comprehensive analysis of the adjacent candidate pairs to obtain the spatial relationship of the adjacent candidate pairs includes: Connect the adjacent candidate pairs to form a three-dimensional straight-line segment of the adjacent candidate pairs; Spatial visibility analysis is performed on the three-dimensional straight line segment to obtain the connectivity information of the three-dimensional straight line segment; Based on the connectivity information, the relational attributes of the adjacent candidate pairs are assigned to obtain the spatial relationship of the adjacent candidate pairs.

7. The method for automated inspection nest path planning of unmanned aerial vehicles (UAVs) as described in claim 1, characterized in that, The step of performing multi-constraint evaluation on the initial navigation network to obtain the constraint weights corresponding to the candidate waypoints, and integrating the candidate waypoints with constraint weights higher than a preset weight threshold into the global optimal waypoint set for the target inspection area, includes: Perform constraint cost mapping on the initial navigation network to obtain the network edge cost value of the initial navigation network; Based on a preset dimensional importance coefficient, the edge values ​​of the network are weighted and fused to obtain the edge constraint weights of the initial navigation network. The calculation formula for the weighted fusion is as follows: ； In the formula, Represents the edges in the initial navigation network Edge constraint weights, Indicates the total number of constraint dimensions. Represents the edge In the The value of network edge computing in each dimension Indicates the first The dimensional importance coefficients of each dimension. This represents the preset cost sensitivity adjustment factor. This represents the preset normalization scaling factor. Represents the natural logarithm function. This represents the natural exponential function. This represents the summation operation; Based on the edge constraint weights, constraint aggregation is performed on the network nodes of the initial navigation network to obtain the point constraint weights of the network nodes. The point constraint weights are used as the constraint weights corresponding to the candidate waypoints; The constraint weights are compared with a preset weight threshold, and the candidate waypoints with constraint weights higher than the preset weight threshold are integrated into the global optimal waypoint set of the target inspection area.

8. The method for automated inspection nest path planning of unmanned aerial vehicles (UAVs) as described in claim 7, characterized in that, The step of performing constraint cost mapping on the initial navigation network to obtain the network edge value of the initial navigation network includes: Spatial feature deconstruction is performed on the network edges of the initial navigation network to obtain multidimensional feature vectors of the network edges; The multidimensional feature vector is mapped to a preset cost transformation rule to obtain the basic cost value of the multidimensional feature vector; The basic value is reduced by data to obtain the network edge value of the initial navigation network.

9. The method for automated inspection nest path planning of unmanned aerial vehicles (UAVs) as described in claim 1, characterized in that, The step of making adjacent waypoints in the globally optimal waypoint set traversable to obtain the navigation path for the target inspection area includes: For adjacent waypoints in the global optimal waypoint set, straight-line segments are constructed to obtain the initial path segments of the adjacent waypoints; Connect the initial path segments with polylines to obtain the turning path trajectory of the initial path segments; The turning path trajectory is smoothed to obtain the navigation path for the target inspection area.

10. A path planning system for automated UAV nest inspection, used to implement the path planning method for automated UAV nest inspection as described in claim 1, the system comprising: The multi-source data fusion module is used to perform multi-source data fusion on the target inspection area to obtain the spatial measurement environment data of the target inspection area. The constraint parsing module is used to parse the airspace constraints in the space measurement environment data and, in combination with the inspection task requirements of the target inspection area, determine the navigation path measurement constraints of the target inspection area. The gridded waypoint identification module is used to perform gridded measurement on the target inspection area based on the navigation path measurement constraints, so as to identify candidate waypoints in the target inspection area; The navigation network construction module is used to analyze the spatial relationships of the candidate waypoints based on their spatial location information in order to construct an initial navigation network for the target inspection area. The multi-constraint evaluation and optimization module is used to perform multi-constraint evaluation on the initial navigation network, obtain the constraint weights corresponding to the candidate waypoints, and integrate the candidate waypoints with constraint weights higher than a preset weight threshold into the global optimal waypoint set of the target inspection area. The waypoint connection path generation module is used to make adjacent waypoints in the global optimal waypoint set traversable to obtain the navigation path of the target inspection area.

Images